Using machine learning models based on cardiac magnetic resonance parameters to predict the prognostic in children with myocarditis

Using machine learning models based on cardiac magnetic resonance parameters to predict the prognostic in children with myocarditis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Using machine learning models based on cardiac magnetic resonance parameters to predict the prognostic in children with myocarditis Dongliang Hu, Manman Cui, Xueke Zhang, Yuanyuan Wu, Yan Liu, Duchang Zhai, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5534455/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 24 May, 2025 Read the published version in BMC Pediatrics → Version 1 posted 9 You are reading this latest preprint version Abstract Objective To develop machine learning (ML) models incorporating explanatory cardiac magnetic resonance (CMR) parameters for predicting the prognosis of myocarditis in pediatric patients. Materials and Methods 77 patients with pediatric myocarditis diagnosed clinically between January 2020 and December 2023 were enrolled retrospectively. All patients were examined by ultrasound, electrocardiogram (ECG), serum biomarkers on admission, and CMR scan to obtain 16 explanatory CMR parameters. All patients underwent follow-up echocardiography and CMR. Patients were divided into two groups according to the occurrence of adverse cardiac events (ACE) during follow-up: the poor prognosis group (n = 23) and the good prognosis group (n = 54). Four models were established, including logistic regression (LR), random forest (RF), support vector machine classifier (SVC), and extreme gradient boosting (XGBoost) model. The performance of each model was evaluated by the area under the receiver operating characteristic curve (AUC). Model interpretation was generated by Shapley additive interpretation. Results Among the four models, the three most important features were late gadolinium enhancement (LGE), left ventricular ejection fraction (LVEF), and SAXPeak Global Circumferential Strain (SAXGCS). In addition, LGE, LVEF, SAXGCS, and LAXPeak Global Longitudinal Strain (LAXGLS) were selected as the key predictors for all four models. Four interpretable CMR parameters were extracted, among which the LR model had the best prediction performance. The AUC, sensitivity, and specificity were 0.893, 0.820, and 0.944, respectively. The findings indicate that the presence of LGE on CMR imaging, along with reductions in LVEF, SAXGCS, and LAXGLS, are predictive of poor prognosis in patients with acute myocarditis. Conclusion ML models, particularly the LR model, demonstrate the potential to predict the prognosis of children with myocarditis. These findings provide valuable insights for cardiologists, supporting more informed clinical decision-making and potentially enhancing patient outcomes in pediatric myocarditis cases. Myocarditis magnetic resonance imaging major cardiovascular adverse events machine learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Myocarditis is an acute or chronic inflammatory disease of the myocardium that can be caused by infectious pathogens such as viruses, bacteria, fungi and Chlamydia, as well as by toxic and hypersensitivity reactions[1]. The pathological features include degeneration, necrosis and fibrosis of myocardial cells, which may eventually cause severe structural and functional impairment of the heart muscle[1]. The increase in the number of patients with myocarditis has been reported after the COVID-19 pandemic. In addition, COVID-19 vaccine-related myocarditis is a condition in which a person experiences localized or diffuse inflammatory changes in the heart muscle as the main manifestation after receiving the COVID-19 vaccine[2]. In the United States, there have been nearly 1,300 reports of myocarditis related to vaccination among over 350 million doses administered, and there have been fatal cases of vaccine-related myocarditis in the United States, Israel, and other places[2]. Acute myocarditis comprises a broad clinical spectrum, from subclinical disease to severe heart failure, and is a major cause of sudden death in young adults. Pathologically, it is characterized by inflammatory cell infiltration of the myocardium with evidence of myocyte necrosis that is not characteristic of an ischemic etiology. Myocarditis can be caused by infections, immune-mediated injury, and toxins (such as anthracyclines)[1]. MRI has high spatial fidelity, good repeatability, strong diagnostic consistency, functional imaging, and quantitative analysis. Cardiovascular magnetic resonance imaging (CMR) is a recognized technique for diagnosing cardiovascular diseases, which is highly specific, sensitive, and non-invasive. It is a very practical method when the clinical diagnosis is unclear, and the specificity of auxiliary examination is not robust. Based on the results of multiple studies [3, 4], the sensitivity of CMR in diagnosing myocarditis is 60–85%, the specificity is 68–90%, and the diagnostic accuracy is close to 80%, which is in line with the diagnostic patterns of myocarditis. It can accurately evaluate the shape and function of the heart, quantify myocardial strain and perfusion function, detect the existence of myocardial edema and fibrosis, and provide important reference value for the diagnosis and prognosis of myocarditis[5–7]. LGE is a marker of myocardial fibrosis and can reflect the existence of myocardial fibrosis[8, 9]. Related studies have confirmed the value of LGE in predicting the prognosis of patients with myocarditis[10–13]. In recent years, imaging techniques have been widely used for clinical risk stratification and prognosis prediction of diseases. CMR imaging is an effective tool for predicting cardiac events[14–16]. However, the traditional features visually extracted from images cannot be fully explained by medical knowledge. In this study, the detailed parameters of cardiac function, strain and tissue characteristics were further extracted by CMR analysis software, and an interpretable prediction model was constructed. Most of the models established by interpretable CMR parameters to predict the prognosis of myocarditis are concentrated in adult patients with myocarditis, while few studies have evaluated the prognosis of children with myocarditis[10–13]. Therefore, this study aimed to explore the feasibility of the machine-learning model based on interpretable CMR parameters for predicting the prognosis of children with myocarditis. Materials And Methods Patient enrollment The study was performed with approvals from the Second Affiliated Hospital of Soochow University institutional board and ethical committee, and was carried out in strict accordance with the relevant guidelines for the acquisition and use of human information and specimens, and the Declaration of Helsinki. Informed consent was waived because of the retrospective nature of the study. Patients who underwent CMR examinations (n = 93) between January 2020 and December 2023 were retrospectively enrolled. Patients without follow-up data (n = 7), with unqualified CMR data (n = 4), and with congenital heart disease (n = 5) were excluded. Ultimately, 77 patients were enrolled in the study. Inclusion criteria and exclusion criteria The inclusion criteria were as follows: (1) age < 18 years; (2) in accordance with the diagnostic guidelines for AHA myocarditis in children in 2022[17]; (3) had undergone selective angiography excluding coronary artery disease; (4) good quality of MRI images. The exclusion criteria were as follows: (1) patients without follow-up evaluation; (2) patients with unqualified CMR data (unqualified CMR data are defined as CMR data with significant motion artifacts, missing key sequences [such as LGE], insufficient image fidelity, or inability to accurately measure myocardial fibrosis and cardiac function parameters); (3) patients with congenital heart disease. Patient characteristics The flowchart of patient enrollment is shown in Fig. 1 . The 77 children included 44 males and 33 females aged 0–16 years, with an average age of 9.4 ± 4.2 years. According to the previous description of myocarditis[18–23], it includes the following manifestations: (1) symptoms and signs of acute myocarditis within 2 weeks of admission, such as fever, prodromal symptoms of the virus, chest tightness, chest pain, dyspnea, palpitation, headache or syncope, a small number of patients have abdominal pain and diarrhea; (2) evidence of structural or functional abnormalities on echocardiography or CMR; (3) abnormal electrocardiogram; (4) increased serum biomarkers, which include cardiac troponin T (cTnT), creatine kinase MB (CK-MB), myoglobin (MYO), and B-type natriuretic peptide (BNP). The mean follow-up duration was 2.8 ± 2.5 years. Patients were divided into a poor prognosis group (n = 23) or a good prognosis group (n = 54) according to the occurrence of ACEs during follow-up. ACE was defined as follows: (1) death or heart transplantation; (2) re-admitted to hospital for heart failure; (3) persistent ventricular arrhythmias; (4) treatment with implantable cardioverter-defibrillator; (5) follow-up MRI or echocardiography showing left ventricular dysfunction and dilated cardiomyopathy. One of these criteria could be defined. Method Scanning methods and parameters Cardiovascular magnetic resonance was performed via 3.0T MRI (Discovery750W, GE Healthcare, Boston, USA [65 patients] and IngeniaCX, Philips Healthcare, Best, Netherlands [12 patients]) and triggered via retrospective ECG gating, and included the following sequences (Table 1 ): (1) heart film: the balanced steady-state free precession sequence was used for scanning; (2) T2WI: three inversion black blood T2WI sequences were used for scanning; (3) late gadolinium enhancement (LGE): after first pass myocardial perfusion imaging, 0.2mmol/kg gadolinium meglumine (6.654g/15ml, Hengrui Pharmaceuticals, Shanghai, China) was injected, and a phase sensitive inversion recovery (PSIR) sequence was used for scanning after 10min. All the scanning sequences capture 3 long-axis images (four-chamber, two-chamber, and three-chamber images). Film sequence and LGE were used to capture all short-axis images from base to apical. T2WI were used to capture three short-axis images: basal, middle, and apical images. Table 1 CMR scanning sequences and parameters. Discovery750W, GE (n = 12) IngeniaCX, Philips (n = 65) Heart film T2WI LGE Heart film T2WI LGE Flip angle (°) 45 107 15 45 90 25 Repetition time (ms) 3.42 1100 6.69 2.84 1200 4.53 Echo time (ms) 1.54 69 3.16 1.42 75 2.20 Slice thickness(mm) 6 6 6 6 6 6 LGE = Late gadolinium enhancement. Image Analysis All the CMR data were post-processed by commercial software CVI42Client (Circle Cardiovascular Imaging, Calgary, Canada). The myocardium was segmented layer by layer in the short-axis view. The epicardium and endocardium contours were delineated using a semi-automatic method. This process combines region-growing, active contour models, and manual adjustments. By removing interference from papillary muscles and the blood pool, and applying smoothing, accurate segmentation results were achieved. [3]. This segmentation method is endorsed by the Society for Cardiovascular Magnetic Resonance. According to the American Heart Association (AHA) segmented method, the myocardium was divided into 17 segments. The generated quantitative CMR parameters had three parts: (1) 11 quantitative parameters related to cardiac function, which including the ejection fraction (EF), end-diastolic volume (EDV), end-systolic volume (ESV), and so on, automatically generated by processing the short-axis film sequence with CVI short-axis 3D module, (2) three layers of short-axis LGE which relate to myocardial fibrosis, generated by processing the LGE sequence with the CVI organization feature module. LGE parameters measure the intensity of myocardial enhancement globally. It was included in the model as categorical variables (presence/absence of enhancement). Segments with a signal intensity ratio above a predefined threshold (typically > 5 standard deviations above the mean signal intensity of normal myocardium) were considered enhanced. (Fig. 2 ). (3) four LV strain-related quantitative parameters, automatically generated by processing long-axis and short-axis cinematographic sequences with the CVI tissue tracking module (Fig. 3 ). These strains include the SAXPeak Global Circumferential Strain (SAXGCS), SAXPeak Global Radial Strain (SAXGRS), LAXPeak Global Longitudinal Strain (LAXGLS) and LAXPeak Global Radial Strain (LAXGRS). A total of 16 CMR parameters are obtained. To evaluate consistency of each feature, we randomly selected 30 cases for repeated segmentation. In this feature subgroup, the viewer 1 repeated the image segmentation twice, and the viewer 2 divided the image independently to evaluate the reproducibility within and between observers. According to the quantitative reproducibility of the intra-group correlation coefficient (ICC), ICC > 0.90 indicates high consistency. Prediction models and evaluation To predict the prognosis of children with myocarditis, we explored four machine learning algorithms: Logistic Regression (LR) for binary classification tasks, Random Forest (RF) as an ensemble learning method based on decision trees, Support Vector Classifier (SVC) as a kernel-based classification method, and Extreme Gradient Boosting (XGBoost) for its scalable and efficient implementation of gradient boosting on decision trees. All the models were developed in Python (version 3.9.12). The LR, RF, SVC, and XGBoost models are implemented using the Python scikit-learn package. Since this is a single-center study, the models obtained from the random segmentation of training and test data may not be generalized. Therefore, ten cross-validations were used to evaluate the performance of the model. Seven performance indicators were recorded in each iteration, including the AUC, sensitivity, specificity, accuracy, F1 score, positive predictive value (PPV) and negative predictive value (NPV). The model was compared in each dataset using average AUC values from ten iterations. The best-performing ML model was selected and measured by its average AUC value. To further validate the models' performance, we conducted bootstrapping analysis to assess their robustness across different metrics. Bootstrapping was performed with 1,000 resampling iterations. In each iteration, 80% of the original dataset was randomly sampled with replacement to form the training subset, while the remaining 20% was retained as the hold-out validation subset. All machine learning models were retrained on the resampled training data, and their performance was evaluated on the corresponding validation subset. The area under the receiver operating characteristic curve (AUC) was computed for each iteration. Final model performance metrics (mean ± standard deviation) and 95% confidence intervals (CIs) were derived using the percentile method. Shapley additive interpretation (Shap) was used to explain the prediction of the ML model and to select the top ten features that have the most significant impact on the prediction. Statistical Analysis The patients were divided into two groups, the good prognosis group and the poor prognosis group, and the variables were compared between the two groups. Independent sample t-test were used for continuous variables with a normal distribution. Variables with non-normal distribution were analyzed via the Mann-Whitney U test. The chi-square (χ2) test or Fisher's exact test was used to categorical variables. Pearson's chi-square test was systematically applied when all expected cell frequencies exceeded 5 with a total sample size ≥ 40, while Fisher's exact test was preferentially employed under conditions where any expected cell frequency fell below 5 or the total sample size was < 40. All the statistical analyses used Python (version 3.9.12) and R software (version 4.2.1), and P < 0.05 was considered significant. Results Cohort characteristics:77 patients with clinical features were divided into two groups. The most common symptoms were chest pain (n = 47) and fever (n = 32), followed by abdominal pain (n = 25) and respiratory symptoms (n = 17). A few patients developed dizziness and headache (n = 14). 7 patients had a recent history of novel coronavirus infection (within 3 months). Most patients had abnormal ECG (77.9%), and 26 patients had ventricular tachycardia (33.8%). The elevation of the ST-segment was the second common (32.5%), followed by depression of the ST-segment (11.7%). 7 patients underwent Endomyocardial biopsy (EMB), of whom 6 were diagnosed with acute myocarditis. In addition, 6 patients had pericardial effusion. There was no significant difference in infection type, clinical manifestation, heart rate, BMI, serum biomarkers and ultrasonography result between the good and poor prognosis groups (Table 2 ). At the end of the follow-up, all patients survived, and no patients received heart transplantation. Table 2 Patient characteristics. ALL Good prognosis Poor prognosis P-value Age 9.36 ± 4.12 9.32 ± 4.04 9.44 ± 4.56 0.907 Female 41 30 11 0.540 Chest pain 47 35 12 0.304 Fever 32 23 9 0.781 Abdominal pain 25 18 7 0.807 Respiratory symptoms 17 12 5 0.963 Headache 14 8 6 0.246 Type of infection 34 20 14 0.055 CRP 3.74(0.89, 13.2) 3.45(0.91,16.80) 4.30(0.10,10.25) 0.269 cTNT 114.60(48.13,446.80) 110.75(41.01,393.10) 134.5(68.84,490.60) 0.302 MYO 30.04(16.0,94.0) 30.07(16.5,87.45) 28.01(17.0,136.56) 0.800 BNP 254.0(90.0,626.2) 206.80(72.60,445.45) 456.0(184.9,1397.0) 0.264 CKMB 9.80(2.70,20.12) 10.35(3.15,21.66) 8.9(2.55,19.15) 0.328 ST elevation 25 10 15 < 0.001 ST depression 9 6 3 0.812 Ventricular tachycardia 26 16 10 0.245 BMI 18.21 ± 3.99 18.49 ± 4.05 17.56 ± 3.86 0.354 BSA 1.16 ± 0.43 1.18 ± 0.44 1.11 ± 0.40 0.511 LVEDV V/B 76.18 ± 14.71 74.07 ± 12.41 81.12 ± 18.44 0.054 LVESV V/B 25.32(21.76,31.32) 25.13(21.85,31.12) 27.01(22.14,31.59) 0.236 LVEF 64.0(55.0, 68.0) 66.0(62.0,69.0) 49.0(46.0,59.0) < 0.001 LVCI 4.44 ± 0.94 4.25 ± 0.88 4.89 ± 0.95 0.005 LVMASS V/B 45.70(40.14,52.18) 44.89(39.59,48.74) 49.72(43.27,61.62) 0.012 RVEDV V/B 69.12 ± 17.05 69.45 ± 15.42 68.35 ± 20.75 0.796 RVESV V/B 28.89 ± 7.66 28.84 ± 6.78 29.01 ± 9.57 0.929 RVEF 57.857 ± 5.33 58.74 ± 4.24 55.78 ± 6.97 0.025 RVCI 3.44 ± 1.09 3.43 ± 0.97 3.47 ± 1.35 0.861 SAXPeak Global Circumferential strain 0.17(0.11,0.20) 0.18(0.16,0.2) 0.10(0.05,0.15) < 0.001 SAXPeak Global Radial strain 0.32 ± 0.08 0.32 ± 0.08 0.31 ± 0.09 0.516 LAXPeak Global Longitudinal strain 0.13 ± 0.04 0.14 ± 0.04 0.09 ± 0.04 < 0.001 LAXPeak Global Radial strain 0.22 ± 0.09 0.22 ± 0.09 0.20 ± 0.08 0.323 LGE 19 2 17 < 0.001 The data are presented as n (%) or mean ± SD. CRP = C-reactive protein cTNT = cardiac troponinT, MYO = myoglobin, BNP = b-type natriuretic peptide, CK-MB = creatine kinase MB, BMI = body mass index, BSA = body surface area, LV = left ventricular, EDV = end-diastolic volume, ESV = end-systolic volume, EF = ejection fraction, CI = cardiac index, RV = right ventricular, LGE = late gadolinium enhancement, V/B = Value/Body surface area. CMR results: Patients with myocarditis were examined by cardiovascular MRI in the hospital within 4.5 ± 7.5 days after onset. Univariate analysis revealed significant differences in LGE, LVEF, SAXGCS, LAXGLS, the LV cardiac index (LVCI), LVEF is the percentage of ejected volume from the left ventricle during systole relative to end-diastolic volume. LVMASS V/B (Value/Body surface area), RVEF, and ST elevation between the two groups, all patients with an average LVEF of 64.0%. Compared with adult myocarditis patients, LVEF damage was less common in children. The average RVEF is 57.9%. 39% of LGE cases were located in the inferior lateral wall, and 42% were in the middle interventricular septum. There was no significant difference in other cardiac function parameters, such as the LVEDV V/B or RVCI, between the patient group and the control group (Table 2 ). Predictive performance of the ML models All 32 features are included in the model input, including 16 clinical features such as demographics, clinical symptoms of myocarditis, laboratory tests, ECG results, and 16 CMR parameters (Table 2 ). Among the ML models considered, the LR model (AUC = 0.893) is superior to the RF (AUC = 0.884), SVC (AUC = 0.880), and XGBOOST (AUC = 0.840) models (Fig. 4 and Table 3 ). The Youden index was used to optimize the index and decision threshold of each model. Under this optimization threshold, the sensitivity and specificity of the LR model are 82.0% and 94.4%, respectively. Bootstrap validation with 1,000 iterations further corroborated the LR model's robustness, demonstrating consistent performance superiority (AUC = 0.895) over RF (AUC = 0.865), SVC (AUC = 0.862), and XGBoost ( AUC = 0.828). The ROC curves for all models, with translucent shading representing 95% confidence intervals (CIs) superimposed, are available in the Supplementary Material 2. Table 3 Performance of each predictive model. LR RF SVC XGBoost Accuracy 0.870 0.857 0.805 0.870 Sensitivity 0.820 0.786 0.736 0.795 Specificity 0.944 0.963 0.907 0.981 AUC 0.893 0.884 0.880 0.840 PPV 0.842 0.875 0.722 0.933 NPV 0.879 0.852 0.831 0.855 F1-score 0.836 0.811 0.751 0.825 Decision thresholds were optimized via the Youden index. LR = logistic regression, RF = random forest, SVC = support vector machine classifier, XGBoost = extreme gradient boosting, AUC = area under the curve, CI = confidence interval, PPV = positive predictive value, NPV = negative predictive value. Model interpretation Shap software was used to explain the prediction of four ML models (LR, SVC, RF, and XGBOOST). Figure 5 shows the functional importance ranking based on the mean | Shap Value |, with the top ten features filtered out for each model. Among all the models, LGE, LVEF, SAXGCS and LAXGLS were the four most important features. Figure 6 shows the positive or negative contributions of the top 10 features to the prognosis of myocarditis in children. In Fig. 6 , each point represents a data sample, and the color indicates whether the observation of the feature itself is greater (redder) or lower (bluer). The features, including LGE, reduced SAXGCS, impaired LAXGLS, decreased LVEF, and ST-segment elevation on ECG, demonstrate a positive correlation with ACE. Discussion In this study, the clinical and CMR data of children with myocarditis were involved to predict the prognosis via ML algorithm, including LR, RF, SVC, and XGBoost models. The prediction effect of the LR model was the best, and the AUC value was the highest. The four ML models screened out important prognostic factors, including LGE, LVEF, SAXGCS, LAXGLS, and ECG ST-segment elevation. Predicting the occurrence of ACE in children with myocarditis is important for early clinical treatment. This study chose LR and nonlinear models (including SVC, RF, and XGBoost) to ensure the reliability and robustness of our results. This approach helps to mitigate potential biases or inaccuracies that might arise from relying on a single model, thereby providing a more balanced and comprehensive evaluation of the data. In the receiver operating characteristic (ROC) analysis, the LR model had the highest AUC value of 0.893 and outperformed the LR, SVC, and XGBoost models. LGE is a marker of myocardial fibrosis and can reflect the existence of myocardial fibrosis[8, 9]. Related studies have confirmed the value of LGE in predicting the prognosis of patients with myocarditis[10–13]. It is reported that the presence of LGE is an independent predictor of poor prognosis, defined as heart transplantation, the need for extracorporeal membrane oxygenation or a ventricular assist device, and death[24, 25], which is consistent with our results. In addition, a recent long-term prognosis study in patients with acute myocarditis revealed that NYHA functional grade II and a larger range of LGE were independent predictors of long-term MACE[26]. However, Aquaro et al.[10, 11]found that LGE in patients with myocarditis mainly existed in the subepicardium of the inferior lateral wall (41%) and anterior middle septum (36%). In contrast, patients with LGE located in the anterior middle septum had a poor prognosis, similar to the results of our study. In our study, LGE was detected among 24.7% of the patients, of which approximately 39% were in the inferior lateral wall and 42% in the middle interventricular septum. Therefore, we believe that the location of LGE has greater prognostic significance than the range of LGE, which is consistent with previous studies[10, 11]. This conclusion has high clinical value and can be used to guide the clinical prediction of myocarditis and to take preventive measures as soon as possible. LVEF refers to the percentage of left ventricular end-systolic ejection volume to left ventricular end-diastolic volume (LVEDV), which can reflect the degree of impaired cardiac function. Some studies have shown that the value of LVEF is lower in patients who died from myocarditis[27]. Patients with an LVEF < 0.30 have a poor prognosis and a significantly increased risk of mechanical circulatory support, death, or heart transplantation (P < 0.001)[27]. Cardiac dysfunction can strongly activate the natriuretic peptide system, and increased ventricular load can lead to the release of B-type natriuretic peptide (BNP). Patients with myocarditis with an LVEF 10000 ng/L) was a risk factor for poor prognosis of pediatric myocarditis[25]. Myocardial strain (MS) refers to the degree of deformation of a cardiac segment from its original shape (end-diastole) to its maximum length (end-systole) in a specified direction and is expressed in percentage terms of the deformation. FT-CMR (feather tracking cardiovascular magnetic resonance imaging) can be used to quantitatively evaluate changes in myocardial strain of patients with myocarditis based on conventional CMR films and analyze the degree and difference in myocardial systolic and diastolic function damage, which is highly valuable for the diagnosis and prognosis of myocarditis. According to a recent review[29], a decreased GCS, or locally circumferential myocardial dysfunction, represents a response to increased wall stress and reflects local changes in myocardial characteristics, such as fibrosis or ischemia caused by microvascular disease or coronary atherosclerotic heart disease (CAD). This increased afterload may lead to progressive myocardial remodeling and the development of dysfunction, leading to a poor prognosis[30]. In addition, there was a significant correlation between the GCS and the LV quality index, which reemphasized the relationship between strain reduction and subclinical heart failure, which may be transformed into symptomatic disease due to poor ventricular remodeling[30, 31]. In addition, studies have shown that the left ventricular strain parameters, especially GLS, are impaired in patients with myocarditis compared with healthy volunteers[32, 33]. GLS is considered to be a strong predictor of major ACE in immune point inhibitor-associated myocarditis[32]. Changes in myocardial strain parameters, especially GLS, are considered to be useful in detecting early changes in cardiac insufficiency[34–36]. Early cardiac dysfunction in most progressive cardiomyopathies leads to a decrease in left ventricular longitudinal mechanics, especially in dilated cardiomyopathy[37]. GLS is an independent predictor of survival in patients with dilated cardiomyopathy[34]. It is speculated that in children with myocarditis, more severe damage to the myocardial short-axis GCS and long-axis GLS strain parameters will increase the risk of left ventricular dysfunction and long-term dilated cardiomyopathy, thus affect the prognosis of patients. ST-segment elevation is the most common change in the ST-segment in acute myocarditis, but ST-segment depression also occurs. In myocarditis, two ST-segment elevation patterns have been described: pericarditis or the typical mode of myocardial infarction. In a study by Nucifora G et al., total ST-segment elevation in all leads was more significant in patients with larger LGE[38]. Our model shows that ECG ST-segment elevation is closely related to prognosis in children with myocarditis. This retrospective study has several limitations. First, the relatively small sample size limits the generalizability of the findings and precludes subgroup analysis. Future studies with larger cohorts and prospective designs are needed to enhance the robustness and persuasiveness of the results. Second, most patients were diagnosed clinically without endocardial biopsy confirmation, which may introduce diagnostic uncertainty. Third, the latest Lake Louise criteria have reduced the diagnostic emphasis on early gadolinium enhancement (EGE), leading to its exclusion from the final analysis. This may affect the comprehensiveness of the imaging assessment. Addressing these limitations in future research could further strengthen the validity and applicability of the findings. Conclusion In summary, ML models, particularly the LR model, which are based on clinical and imaging data from pediatric myocarditis patients, can effectively predict the prognosis of children with myocarditis. The optimal ML model (LR) offers early warning capabilities and supports more informed treatment strategies. This study will further investigate laboratory and imaging data related to pediatric myocarditis to refine the models and achieve greater diagnostic accuracy. Declarations Ethics approval and consent to participate: The study was performed with approvals from the Second Affiliated Hospital of Soochow University institutional board and ethical committee, and was carried out in strict accordance with the relevant guidelines for the acquisition and use of human information and specimens, and the Declaration of Helsinki. Informed consent was waived because of the retrospective nature of the study. Clinical trial number: Not applicable. Author Contributions: Conceptualization: Dongliang Hu, Manman Cui. Data curation: Xueke Zhang, Yuanyuan Wu. Formal analysis: Yan Liu，Duchang Zhai. Funding acquisition: Guohua Fan, Wu Cai. Investigation: Wanliang Guo. Methodology: Dongliang Hu. Project administration: Guohua Fan ，Wu Cai. Resources: Wanliang Guo，Wu Cai. Software: Manman Cui. Supervision: Shenghong Ju, Wu Cai. Validation: Dongliang Hu, Manman Cui. Visualization: Dongliang Hu, Manman Cui. Writing—original draft: Dongliang Hu, Manman Cui. Writing—review & editing: all authors. Consent for publication: The authors have agreed to publish this article. Conflicts of interest statement: The authors have no conflicts of interest with respect to this study. Availability of data and materials: The datasets generated or analyzed during the study are available from the corresponding author on reasonable request. Funding: This study was supported by the Project of State Key Laboratory of Radiation Medicine and Protection, Soochow University (GZK1202136). References Sagar S, Liu PP, Cooper LT, Jr. Myocarditis. Lancet (London, England).2012; 379(9817):738-747. http://doi.org/10.1016/s0140-6736(11)60648-x Shiravi AA, Ardekani A, Sheikhbahaei E, Heshmat-Ghahdarijani K. Cardiovascular Complications of SARS-CoV-2 Vaccines: An Overview. Cardiology and therapy.2022; 11(1):13-21. http://doi.org/10.1007/s40119-021-00248-0 Lee JW, Jeong YJ, Lee G, Lee NK, Lee HW, Kim JY, Choi BS, Choo KS. Predictive Value of Cardiac Magnetic Resonance Imaging-Derived Myocardial Strain for Poor Outcomes in Patients with Acute Myocarditis. Korean journal of radiology.2017; 18(4):643-654. http://doi.org/10.3348/kjr.2017.18.4.643 Weinreich MA, Jabbar AY, Malguria N, Haley RW. New-Onset Myocarditis in an Immunocompetent Adult with Acute Metapneumovirus Infection. Case reports in medicine.2015; 2015:814269. http://doi.org/10.1155/2015/814269 Ferreira VM, Schulz-Menger J, Holmvang G, Kramer CM, Carbone I, Sechtem U, Kindermann I, Gutberlet M, Cooper LT, Liu P et al . Cardiovascular Magnetic Resonance in Nonischemic Myocardial Inflammation: Expert Recommendations. Journal of the American College of Cardiology.2018; 72(24):3158-3176. http://doi.org/10.1016/j.jacc.2018.09.072 Wang H, Zhao B, Jia H, Gao F, Zhao J, Wang C. A retrospective study: cardiac MRI of fulminant myocarditis in children-can we evaluate the short-term outcomes? PeerJ.2016; 4:e2750. http://doi.org/10.7717/peerj.2750 Di Filippo S. Improving outcomes of acute myocarditis in children. Expert review of cardiovascular therapy.2016; 14(1):117-125. http://doi.org/10.1586/14779072.2016.1114884 Messroghli DR, Moon JC, Ferreira VM, Grosse-Wortmann L, He T, Kellman P, Mascherbauer J, Nezafat R, Salerno M, Schelbert EB et al . Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: A consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR) endorsed by the European Association for Cardiovascular Imaging (EACVI). Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance.2017; 19(1):75. http://doi.org/10.1186/s12968-017-0389-8 Florian A, Ludwig A, Rösch S, Yildiz H, Sechtem U, Yilmaz A. Myocardial fibrosis imaging based on T1-mapping and extracellular volume fraction (ECV) measurement in muscular dystrophy patients: diagnostic value compared with conventional late gadolinium enhancement (LGE) imaging. European heart journal Cardiovascular Imaging.2014; 15(9):1004-1012. http://doi.org/10.1093/ehjci/jeu050 Aquaro GD, Ghebru Habtemicael Y, Camastra G, Monti L, Dellegrottaglie S, Moro C, Lanzillo C, Scatteia A, Di Roma M, Pontone G et al . Prognostic Value of Repeating Cardiac Magnetic Resonance in Patients With Acute Myocarditis. Journal of the American College of Cardiology.2019; 74(20):2439-2448. http://doi.org/10.1016/j.jacc.2019.08.1061 Aquaro GD, Perfetti M, Camastra G, Monti L, Dellegrottaglie S, Moro C, Pepe A, Todiere G, Lanzillo C, Scatteia A et al . Cardiac MR With Late Gadolinium Enhancement in Acute Myocarditis With Preserved Systolic Function: ITAMY Study. Journal of the American College of Cardiology.2017; 70(16):1977-1987. http://doi.org/10.1016/j.jacc.2017.08.044 Blissett S, Chocron Y, Kovacina B, Afilalo J. Diagnostic and prognostic value of cardiac magnetic resonance in acute myocarditis: a systematic review and meta-analysis. The international journal of cardiovascular imaging.2019; 35(12):2221-2229. http://doi.org/10.1007/s10554-019-01674-x Yang F, Wang J, Li W, Xu Y, Wan K, Zeng R, Chen Y. The prognostic value of late gadolinium enhancement in myocarditis and clinically suspected myocarditis: systematic review and meta-analysis. European radiology.2020; 30(5):2616-2626. http://doi.org/10.1007/s00330-019-06643-5 Leiner T, Rueckert D, Suinesiaputra A, Baeßler B, Nezafat R, Išgum I, Young AA. Machine learning in cardiovascular magnetic resonance: basic concepts and applications. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance.2019; 21(1):61. http://doi.org/10.1186/s12968-019-0575-y Zhang N, Yang G, Gao Z, Xu C, Zhang Y, Shi R, Keegan J, Xu L, Zhang H, Fan Z et al . Deep Learning for Diagnosis of Chronic Myocardial Infarction on Nonenhanced Cardiac Cine MRI. Radiology.2019; 291(3):606-617. http://doi.org/10.1148/radiol.2019182304 Baessler B, Mannil M, Oebel S, Maintz D, Alkadhi H, Manka R. Subacute and Chronic Left Ventricular Myocardial Scar: Accuracy of Texture Analysis on Nonenhanced Cine MR Images. Radiology.2018; 286(1):103-112. http://doi.org/10.1148/radiol.2017170213 Law YM, Lal AK, Chen S, Čiháková D, Cooper LT, Jr., Deshpande S, Godown J, Grosse-Wortmann L, Robinson JD, Towbin JA. Diagnosis and Management of Myocarditis in Children: A Scientific Statement From the American Heart Association. Circulation.2021; 144(6):e123-e135. http://doi.org/10.1161/cir.0000000000001001 Hsiao JF, Koshino Y, Bonnichsen CR, Yu Y, Miller FA, Jr., Pellikka PA, Cooper LT, Jr., Villarraga HR. Speckle tracking echocardiography in acute myocarditis. The international journal of cardiovascular imaging.2013; 29(2):275-284. http://doi.org/10.1007/s10554-012-0085-6 Friedrich MG, Sechtem U, Schulz-Menger J, Holmvang G, Alakija P, Cooper LT, White JA, Abdel-Aty H, Gutberlet M, Prasad S et al . Cardiovascular magnetic resonance in myocarditis: A JACC White Paper. Journal of the American College of Cardiology.2009; 53(17):1475-1487. http://doi.org/10.1016/j.jacc.2009.02.007 Schultz JC, Hilliard AA, Cooper LT, Jr., Rihal CS. Diagnosis and treatment of viral myocarditis. Mayo Clinic proceedings.2009; 84(11):1001-1009. http://doi.org/10.1016/s0025-6196(11)60670-8 Mahrholdt H, Goedecke C, Wagner A, Meinhardt G, Athanasiadis A, Vogelsberg H, Fritz P, Klingel K, Kandolf R, Sechtem U. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation.2004; 109(10):1250-1258. http://doi.org/10.1161/01.Cir.0000118493.13323.81 Caforio AL, Pankuweit S, Arbustini E, Basso C, Gimeno-Blanes J, Felix SB, Fu M, Heliö T, Heymans S, Jahns R et al . Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: a position statement of the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. European heart journal.2013; 34(33):2636-2648, 2648a-2648d. http://doi.org/10.1093/eurheartj/eht210 Magnani JW, Dec GW. Myocarditis: current trends in diagnosis and treatment. Circulation.2006; 113(6):876-890. http://doi.org/10.1161/circulationaha.105.584532 Grün S, Schumm J, Greulich S, Wagner A, Schneider S, Bruder O, Kispert EM, Hill S, Ong P, Klingel K et al . Long-term follow-up of biopsy-proven viral myocarditis: predictors of mortality and incomplete recovery. Journal of the American College of Cardiology.2012; 59(18):1604-1615. http://doi.org/10.1016/j.jacc.2012.01.007 Sachdeva S, Song X, Dham N, Heath DM, DeBiasi RL. Analysis of clinical parameters and cardiac magnetic resonance imaging as predictors of outcome in pediatric myocarditis. The American journal of cardiology.2015; 115(4):499-504. http://doi.org/10.1016/j.amjcard.2014.11.029 André F, Stock FT, Riffel J, Giannitsis E, Steen H, Scharhag J, Katus HA, Buss SJ. Incremental value of cardiac deformation analysis in acute myocarditis: a cardiovascular magnetic resonance imaging study. The international journal of cardiovascular imaging.2016; 32(7):1093-1101. http://doi.org/10.1007/s10554-016-0878-0 Schubert S, Opgen-Rhein B, Boehne M, Weigelt A, Wagner R, Müller G, Rentzsch A, Zu Knyphausen E, Fischer M, Papakostas K et al . Severe heart failure and the need for mechanical circulatory support and heart transplantation in pediatric patients with myocarditis: Results from the prospective multicenter registry "MYKKE". Pediatric transplantation.2019; 23(7):e13548. http://doi.org/10.1111/petr.13548 Akgül F, Er A, Ulusoy E, Çağlar A, Vuran G, Seven P, Yılmazer MM, Ağın H, Apa H. Are clinical features and cardiac biomarkers at admission related to severity in pediatric acute myocarditis?: Clinical features and cardiac biomarkers in pediatric acute myocarditis. Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.2022; 29(5):376-380. http://doi.org/10.1016/j.arcped.2022.03.008 Korosoglou G, Sagris M, André F, Steen H, Montenbruck M, Frey N, Kelle S. Systematic review and meta-analysis for the value of cardiac magnetic resonance strain to predict cardiac outcomes. Scientific reports.2024; 14(1):1094. http://doi.org/10.1038/s41598-023-50835-5 Kass DA. Ventricular arterial stiffening: integrating the pathophysiology. (1524-4563 (Electronic)). Rosen BD, Edvardsen T Fau - Lai S, Lai S Fau - Castillo E, Castillo E Fau - Pan L, Pan L Fau - Jerosch-Herold M, Jerosch-Herold M Fau - Sinha S, Sinha S Fau - Kronmal R, Kronmal R Fau - Arnett D, Arnett D Fau - Crouse JR, 3rd, Crouse Jr 3rd Fau - Heckbert SR et al . Left ventricular concentric remodeling is associated with decreased global and regional systolic function: the Multi-Ethnic Study of Atherosclerosis. (1524-4539 (Electronic)). Awadalla M, Mahmood SS, Groarke JD, Hassan MZO, Nohria A, Rokicki A, Murphy SP, Mercaldo ND, Zhang L, Zlotoff DA et al . Global Longitudinal Strain and Cardiac Events in Patients With Immune Checkpoint Inhibitor-Related Myocarditis. Journal of the American College of Cardiology.2020; 75(5):467-478. http://doi.org/10.1016/j.jacc.2019.11.049 Luetkens JA, Schlesinger-Irsch U, Kuetting DL, Dabir D, Homsi R, Doerner J, Schmeel FC, Fimmers R, Sprinkart AM, Naehle CP et al . Feature-tracking myocardial strain analysis in acute myocarditis: diagnostic value and association with myocardial oedema. European radiology.2017; 27(11):4661-4671. http://doi.org/10.1007/s00330-017-4854-4 Buss SJ, Breuninger K, Lehrke S, Voss A, Galuschky C, Lossnitzer D, Andre F, Ehlermann P, Franke J, Taeger T et al . Assessment of myocardial deformation with cardiac magnetic resonance strain imaging improves risk stratification in patients with dilated cardiomyopathy. European heart journal Cardiovascular Imaging.2015; 16(3):307-315. http://doi.org/10.1093/ehjci/jeu181 Pedrizzetti G, Claus P, Kilner PJ, Nagel E. Principles of cardiovascular magnetic resonance feature tracking and echocardiographic speckle tracking for informed clinical use. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance.2016; 18(1):51. http://doi.org/10.1186/s12968-016-0269-7 Amzulescu MS, De Craene M, Langet H, Pasquet A, Vancraeynest D, Pouleur AC, Vanoverschelde JL, Gerber BL. Myocardial strain imaging: review of general principles, validation, and sources of discrepancies. European heart journal Cardiovascular Imaging.2019; 20(6):605-619. http://doi.org/10.1093/ehjci/jez041 Claus P, Omar AMS, Pedrizzetti G, Sengupta PP, Nagel E. Tissue Tracking Technology for Assessing Cardiac Mechanics: Principles, Normal Values, and Clinical Applications. JACC Cardiovascular imaging.2015; 8(12):1444-1460. http://doi.org/10.1016/j.jcmg.2015.11.001 Nucifora G, Miani D, Di Chiara A, Piccoli G, Artico J, Puppato M, Slavich G, De Biasio M, Gasparini D, Proclemer A. Infarct-like acute myocarditis: relation between electrocardiographic findings and myocardial damage as assessed by cardiac magnetic resonance imaging. Clinical cardiology.2013; 36(3):146-152. http://doi.org/10.1002/clc.22088 Additional Declarations No competing interests reported. Supplementary Files supplementarymaterial.docx supplementarymaterial2.docx Cite Share Download PDF Status: Published Journal Publication published 24 May, 2025 Read the published version in BMC Pediatrics → Version 1 posted Editorial decision: Revision requested 02 May, 2025 Editor assigned by journal 02 May, 2025 Reviews received at journal 28 Apr, 2025 Reviewers agreed at journal 22 Apr, 2025 Reviews received at journal 22 Apr, 2025 Reviewers agreed at journal 22 Apr, 2025 Reviewers invited by journal 22 Apr, 2025 Submission checks completed at journal 18 Apr, 2025 First submitted to journal 16 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5534455","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":446566796,"identity":"5f9f39e0-f3c1-4654-b506-a1b6bafd2baa","order_by":0,"name":"Dongliang Hu","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Dongliang","middleName":"","lastName":"Hu","suffix":""},{"id":446566797,"identity":"43aeeb46-10b1-40d2-9053-4a6a66612577","order_by":1,"name":"Manman Cui","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Manman","middleName":"","lastName":"Cui","suffix":""},{"id":446566798,"identity":"408e200d-3c3d-4c77-839c-654ef795262d","order_by":2,"name":"Xueke Zhang","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Xueke","middleName":"","lastName":"Zhang","suffix":""},{"id":446566799,"identity":"d05bf427-330d-463d-80bd-0677705f546a","order_by":3,"name":"Yuanyuan Wu","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Yuanyuan","middleName":"","lastName":"Wu","suffix":""},{"id":446566800,"identity":"8b06b25a-ead0-4a61-b823-1bf1c8a7434c","order_by":4,"name":"Yan Liu","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Liu","suffix":""},{"id":446566801,"identity":"83dce399-5819-4d62-a87f-3f2725b4a635","order_by":5,"name":"Duchang Zhai","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Duchang","middleName":"","lastName":"Zhai","suffix":""},{"id":446566802,"identity":"e40bdbeb-1031-4dd4-a2db-c848bcab6fc5","order_by":6,"name":"Wanliang Guo","email":"","orcid":"","institution":"Children's Hospital of Suzhou University","correspondingAuthor":false,"prefix":"","firstName":"Wanliang","middleName":"","lastName":"Guo","suffix":""},{"id":446566803,"identity":"20731d2f-65d2-4d40-8174-e4d1aaaa6e2b","order_by":7,"name":"Shenghong Ju","email":"","orcid":"","institution":"Zhongda Hospital Southeast University","correspondingAuthor":false,"prefix":"","firstName":"Shenghong","middleName":"","lastName":"Ju","suffix":""},{"id":446566804,"identity":"bd0ba123-ccd1-4fdd-a633-6098a031eda5","order_by":8,"name":"Guohua Fan","email":"","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Guohua","middleName":"","lastName":"Fan","suffix":""},{"id":446566805,"identity":"e09a65e8-3497-4df5-9bf8-5fe127fe60ad","order_by":9,"name":"Wu Cai","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvklEQVRIiWNgGAWjYHACxgcJFTY8/OwNxGthNvhwJk1GsucA8VrYBGe2HbYxuOFApHqDG+nXmHnOnOdhuMHA+OFjDhFaJGfklD3mqbjNwzi7gVly5jYitPBL56Qb85y5zcMsc4CNmZcYLWzSOWnSvG3neNgkEojUwi+dfkxyZtsBHh6itUjOfwMK5GQeCZ6DzcT5xeDM8YfAqLSztz/efPDDR2K0MDDwGEAZjA1EqQcC9gfEqhwFo2AUjIKRCgA6qjaukBUtSQAAAABJRU5ErkJggg==","orcid":"","institution":"Second Affiliated Hospital of Soochow University","correspondingAuthor":true,"prefix":"","firstName":"Wu","middleName":"","lastName":"Cai","suffix":""}],"badges":[],"createdAt":"2024-11-27 10:23:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5534455/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5534455/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12887-025-05753-y","type":"published","date":"2025-05-24T15:57:12+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81506944,"identity":"a65b2aea-a787-4947-af01-9c6d82a37a12","added_by":"auto","created_at":"2025-04-28 05:32:53","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2232133,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of patient enrollment.\u003c/p\u003e","description":"","filename":"fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/072c981e5334563d36ca6ba1.jpg"},{"id":81506917,"identity":"d07203de-fc5b-4ce5-b7fe-eb27dbe4b038","added_by":"auto","created_at":"2025-04-28 05:32:52","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1444731,"visible":true,"origin":"","legend":"\u003cp\u003eLate gadolinium-enhanced images of the left ventricle short axis; red: endocardial line of the left ventricle; green: left ventricular epicardial line; yellow: delayed reinforcement area (the myocardial gray threshold is 5 standard deviations greater than the average signal strength of normal myocardium to determine late enhancement).\u003c/p\u003e","description":"","filename":"fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/5d10b1eafc74c741216f6439.jpg"},{"id":81506929,"identity":"ce477a6f-4581-470e-b52f-9f7a6c344cbe","added_by":"auto","created_at":"2025-04-28 05:32:52","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2912899,"visible":true,"origin":"","legend":"\u003cp\u003eMyocardial strain measurement via the feature tracking method in an 8-year-old female patient with acute myocarditis. After the endocardial and epicardial borders of the LV were traced semi-automatically, The software (CVI42) automatically tracked the endocardial and epicardial borders across frames during the cardiac cycle. SAXGCS and SAXGRS measurements (A) were obtained via mid-ventricular level short-axis cine views. LAXGRS and LAXGCS measurements (B) were obtained from 2-chamber long-axis view. SAXGCS= SAXPeak Global Circumferential Strain measured from short-axis cine views, LAXGLS= LAXPeak Global Longitudinal Strain measured from long-axis cine views, LAXGRS= LAXPeak Global Radial Strain measured from long-axis cine views, SAXGRS= SAXPeak Global Radial Strain measured from short-axis cine views, LV=left ventricular.\u003c/p\u003e","description":"","filename":"fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/215ca74a508d5d503c62eed1.jpg"},{"id":81506951,"identity":"e4ac63b4-2c2f-4bbe-82bd-6c7395d5b17d","added_by":"auto","created_at":"2025-04-28 05:32:54","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":142804,"visible":true,"origin":"","legend":"\u003cp\u003eAreas under the receiver operating characteristic curve (AUC) for all four machine learning models [logistic regression (LR), random forest (RF), support vector machine classifier (SVC), and extreme gradient boosting (XGBoost)].\u003c/p\u003e","description":"","filename":"fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/146c49a0551bb2ac1856ee06.jpg"},{"id":81506922,"identity":"d38bfc00-c6ba-4b43-8362-2c7621a27386","added_by":"auto","created_at":"2025-04-28 05:32:52","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":356964,"visible":true,"origin":"","legend":"\u003cp\u003eImportance of the top ten features for each model based on Shapley Additive Explanations (SHAP).\u003c/p\u003e","description":"","filename":"fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/5f3af388a8b4f0dfbf9fedb8.jpg"},{"id":81509721,"identity":"9a05a4a5-bee6-49bc-a842-8615b24a5bba","added_by":"auto","created_at":"2025-04-28 06:02:29","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":362085,"visible":true,"origin":"","legend":"\u003cp\u003eShapley Additive Explanations (SHAP) for interpreting the top ten features in each machine learning model.\u003c/p\u003e","description":"","filename":"fig6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/5657ffda9c877cd082940fcd.jpg"},{"id":83459972,"identity":"f50eb4ef-d428-48d3-a4f4-6d065f50a3bc","added_by":"auto","created_at":"2025-05-26 16:07:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8317948,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/c87dac27-31d9-4221-8592-fba8f5e304cf.pdf"},{"id":81508873,"identity":"c8dd2bd2-4d92-440b-af84-4ed7658882c5","added_by":"auto","created_at":"2025-04-28 05:40:52","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":19389,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/c2602aa16e911141a794a084.docx"},{"id":81508892,"identity":"697cbad0-4174-44ae-afb4-59cee39a2797","added_by":"auto","created_at":"2025-04-28 05:40:53","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":148387,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarymaterial2.docx","url":"https://assets-eu.researchsquare.com/files/rs-5534455/v1/99f630e2fa6a2f406f962d46.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Using machine learning models based on cardiac magnetic resonance parameters to predict the prognostic in children with myocarditis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eMyocarditis is an acute or chronic inflammatory disease of the myocardium that can be caused by infectious pathogens such as viruses, bacteria, fungi and Chlamydia, as well as by toxic and hypersensitivity reactions[1]. The pathological features include degeneration, necrosis and fibrosis of myocardial cells, which may eventually cause severe structural and functional impairment of the heart muscle[1]. The increase in the number of patients with myocarditis has been reported after the COVID-19 pandemic. In addition, COVID-19 vaccine-related myocarditis is a condition in which a person experiences localized or diffuse inflammatory changes in the heart muscle as the main manifestation after receiving the COVID-19 vaccine[2]. In the United States, there have been nearly 1,300 reports of myocarditis related to vaccination among over 350\u0026nbsp;million doses administered, and there have been fatal cases of vaccine-related myocarditis in the United States, Israel, and other places[2]. Acute myocarditis comprises a broad clinical spectrum, from subclinical disease to severe heart failure, and is a major cause of sudden death in young adults. Pathologically, it is characterized by inflammatory cell infiltration of the myocardium with evidence of myocyte necrosis that is not characteristic of an ischemic etiology. Myocarditis can be caused by infections, immune-mediated injury, and toxins (such as anthracyclines)[1].\u003c/p\u003e \u003cp\u003eMRI has high spatial fidelity, good repeatability, strong diagnostic consistency, functional imaging, and quantitative analysis. Cardiovascular magnetic resonance imaging (CMR) is a recognized technique for diagnosing cardiovascular diseases, which is highly specific, sensitive, and non-invasive. It is a very practical method when the clinical diagnosis is unclear, and the specificity of auxiliary examination is not robust. Based on the results of multiple studies [3, 4], the sensitivity of CMR in diagnosing myocarditis is 60\u0026ndash;85%, the specificity is 68\u0026ndash;90%, and the diagnostic accuracy is close to 80%, which is in line with the diagnostic patterns of myocarditis. It can accurately evaluate the shape and function of the heart, quantify myocardial strain and perfusion function, detect the existence of myocardial edema and fibrosis, and provide important reference value for the diagnosis and prognosis of myocarditis[5\u0026ndash;7]. LGE is a marker of myocardial fibrosis and can reflect the existence of myocardial fibrosis[8, 9]. Related studies have confirmed the value of LGE in predicting the prognosis of patients with myocarditis[10\u0026ndash;13].\u003c/p\u003e \u003cp\u003eIn recent years, imaging techniques have been widely used for clinical risk stratification and prognosis prediction of diseases. CMR imaging is an effective tool for predicting cardiac events[14\u0026ndash;16]. However, the traditional features visually extracted from images cannot be fully explained by medical knowledge. In this study, the detailed parameters of cardiac function, strain and tissue characteristics were further extracted by CMR analysis software, and an interpretable prediction model was constructed. Most of the models established by interpretable CMR parameters to predict the prognosis of myocarditis are concentrated in adult patients with myocarditis, while few studies have evaluated the prognosis of children with myocarditis[10\u0026ndash;13]. Therefore, this study aimed to explore the feasibility of the machine-learning model based on interpretable CMR parameters for predicting the prognosis of children with myocarditis.\u003c/p\u003e"},{"header":"Materials And Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatient enrollment\u003c/h2\u003e \u003cp\u003e The study was performed with approvals from the Second Affiliated Hospital of Soochow University institutional board and ethical committee, and was carried out in strict accordance with the relevant guidelines for the acquisition and use of human information and specimens, and the Declaration of Helsinki. Informed consent was waived because of the retrospective nature of the study. Patients who underwent CMR examinations (n\u0026thinsp;=\u0026thinsp;93) between January 2020 and December 2023 were retrospectively enrolled. Patients without follow-up data (n\u0026thinsp;=\u0026thinsp;7), with unqualified CMR data (n\u0026thinsp;=\u0026thinsp;4), and with congenital heart disease (n\u0026thinsp;=\u0026thinsp;5) were excluded. Ultimately, 77 patients were enrolled in the study.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eInclusion criteria and exclusion criteria\u003c/h3\u003e\n\u003cp\u003eThe inclusion criteria were as follows: (1) age\u0026thinsp;\u0026lt;\u0026thinsp;18 years; (2) in accordance with the diagnostic guidelines for AHA myocarditis in children in 2022[17]; (3) had undergone selective angiography excluding coronary artery disease; (4) good quality of MRI images.\u003c/p\u003e \u003cp\u003eThe exclusion criteria were as follows: (1) patients without follow-up evaluation; (2) patients with unqualified CMR data (unqualified CMR data are defined as CMR data with significant motion artifacts, missing key sequences [such as LGE], insufficient image fidelity, or inability to accurately measure myocardial fibrosis and cardiac function parameters); (3) patients with congenital heart disease.\u003c/p\u003e\n\u003ch3\u003ePatient characteristics\u003c/h3\u003e\n\u003cp\u003eThe flowchart of patient enrollment is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The 77 children included 44 males and 33 females aged 0\u0026ndash;16 years, with an average age of 9.4\u0026thinsp;\u0026plusmn;\u0026thinsp;4.2 years. According to the previous description of myocarditis[18\u0026ndash;23], it includes the following manifestations: (1) symptoms and signs of acute myocarditis within 2 weeks of admission, such as fever, prodromal symptoms of the virus, chest tightness, chest pain, dyspnea, palpitation, headache or syncope, a small number of patients have abdominal pain and diarrhea; (2) evidence of structural or functional abnormalities on echocardiography or CMR; (3) abnormal electrocardiogram; (4) increased serum biomarkers, which include cardiac troponin T (cTnT), creatine kinase MB (CK-MB), myoglobin (MYO), and B-type natriuretic peptide (BNP). The mean follow-up duration was 2.8\u0026thinsp;\u0026plusmn;\u0026thinsp;2.5 years. Patients were divided into a poor prognosis group (n\u0026thinsp;=\u0026thinsp;23) or a good prognosis group (n\u0026thinsp;=\u0026thinsp;54) according to the occurrence of ACEs during follow-up. ACE was defined as follows: (1) death or heart transplantation; (2) re-admitted to hospital for heart failure; (3) persistent ventricular arrhythmias; (4) treatment with implantable cardioverter-defibrillator; (5) follow-up MRI or echocardiography showing left ventricular dysfunction and dilated cardiomyopathy. One of these criteria could be defined.\u003c/p\u003e\n\u003ch3\u003eMethod\u003c/h3\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eScanning methods and parameters\u003c/h2\u003e \u003cp\u003eCardiovascular magnetic resonance was performed via 3.0T MRI (Discovery750W, GE Healthcare, Boston, USA [65 patients] and IngeniaCX, Philips Healthcare, Best, Netherlands [12 patients]) and triggered via retrospective ECG gating, and included the following sequences (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e): (1) heart film: the balanced steady-state free precession sequence was used for scanning; (2) T2WI: three inversion black blood T2WI sequences were used for scanning; (3) late gadolinium enhancement (LGE): after first pass myocardial perfusion imaging, 0.2mmol/kg gadolinium meglumine (6.654g/15ml, Hengrui Pharmaceuticals, Shanghai, China) was injected, and a phase sensitive inversion recovery (PSIR) sequence was used for scanning after 10min. All the scanning sequences capture 3 long-axis images (four-chamber, two-chamber, and three-chamber images). Film sequence and LGE were used to capture all short-axis images from base to apical. T2WI were used to capture three short-axis images: basal, middle, and apical images.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCMR scanning sequences and parameters.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003eDiscovery750W, GE (n\u0026thinsp;=\u0026thinsp;12)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c7\" namest=\"c5\"\u003e \u003cp\u003eIngeniaCX, Philips (n\u0026thinsp;=\u0026thinsp;65)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHeart film\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eT2WI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLGE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHeart film\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eT2WI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eLGE\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFlip angle (\u0026deg;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e107\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRepetition time (ms)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4.53\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEcho time (ms)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSlice thickness(mm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003eLGE\u0026thinsp;=\u0026thinsp;Late gadolinium enhancement.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eImage Analysis\u003c/h2\u003e \u003cp\u003eAll the CMR data were post-processed by commercial software CVI42Client (Circle Cardiovascular Imaging, Calgary, Canada). The myocardium was segmented layer by layer in the short-axis view. The epicardium and endocardium contours were delineated using a semi-automatic method. This process combines region-growing, active contour models, and manual adjustments. By removing interference from papillary muscles and the blood pool, and applying smoothing, accurate segmentation results were achieved. [3]. This segmentation method is endorsed by the Society for Cardiovascular Magnetic Resonance. According to the American Heart Association (AHA) segmented method, the myocardium was divided into 17 segments. The generated quantitative CMR parameters had three parts: (1) 11 quantitative parameters related to cardiac function, which including the ejection fraction (EF), end-diastolic volume (EDV), end-systolic volume (ESV), and so on, automatically generated by processing the short-axis film sequence with CVI short-axis 3D module, (2) three layers of short-axis LGE which relate to myocardial fibrosis, generated by processing the LGE sequence with the CVI organization feature module. LGE parameters measure the intensity of myocardial enhancement globally. It was included in the model as categorical variables (presence/absence of enhancement). Segments with a signal intensity ratio above a predefined threshold (typically\u0026thinsp;\u0026gt;\u0026thinsp;5 standard deviations above the mean signal intensity of normal myocardium) were considered enhanced. (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). (3) four LV strain-related quantitative parameters, automatically generated by processing long-axis and short-axis cinematographic sequences with the CVI tissue tracking module (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). These strains include the SAXPeak Global Circumferential Strain (SAXGCS), SAXPeak Global Radial Strain (SAXGRS), LAXPeak Global Longitudinal Strain (LAXGLS) and LAXPeak Global Radial Strain (LAXGRS). A total of 16 CMR parameters are obtained. To evaluate consistency of each feature, we randomly selected 30 cases for repeated segmentation. In this feature subgroup, the viewer 1 repeated the image segmentation twice, and the viewer 2 divided the image independently to evaluate the reproducibility within and between observers. According to the quantitative reproducibility of the intra-group correlation coefficient (ICC), ICC\u0026thinsp;\u0026gt;\u0026thinsp;0.90 indicates high consistency.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003ePrediction models and evaluation\u003c/h3\u003e\n\u003cp\u003eTo predict the prognosis of children with myocarditis, we explored four machine learning algorithms: Logistic Regression (LR) for binary classification tasks, Random Forest (RF) as an ensemble learning method based on decision trees, Support Vector Classifier (SVC) as a kernel-based classification method, and Extreme Gradient Boosting (XGBoost) for its scalable and efficient implementation of gradient boosting on decision trees. All the models were developed in Python (version 3.9.12). The LR, RF, SVC, and XGBoost models are implemented using the Python scikit-learn package. Since this is a single-center study, the models obtained from the random segmentation of training and test data may not be generalized. Therefore, ten cross-validations were used to evaluate the performance of the model. Seven performance indicators were recorded in each iteration, including the AUC, sensitivity, specificity, accuracy, F1 score, positive predictive value (PPV) and negative predictive value (NPV). The model was compared in each dataset using average AUC values from ten iterations. The best-performing ML model was selected and measured by its average AUC value. To further validate the models' performance, we conducted bootstrapping analysis to assess their robustness across different metrics. Bootstrapping was performed with 1,000 resampling iterations. In each iteration, 80% of the original dataset was randomly sampled with replacement to form the training subset, while the remaining 20% was retained as the hold-out validation subset. All machine learning models were retrained on the resampled training data, and their performance was evaluated on the corresponding validation subset. The area under the receiver operating characteristic curve (AUC) was computed for each iteration. Final model performance metrics (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation) and 95% confidence intervals (CIs) were derived using the percentile method. Shapley additive interpretation (Shap) was used to explain the prediction of the ML model and to select the top ten features that have the most significant impact on the prediction.\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eThe patients were divided into two groups, the good prognosis group and the poor prognosis group, and the variables were compared between the two groups. Independent sample t-test were used for continuous variables with a normal distribution. Variables with non-normal distribution were analyzed via the Mann-Whitney U test. The chi-square (χ2) test or Fisher's exact test was used to categorical variables. Pearson's chi-square test was systematically applied when all expected cell frequencies exceeded 5 with a total sample size\u0026thinsp;\u0026ge;\u0026thinsp;40, while Fisher's exact test was preferentially employed under conditions where any expected cell frequency fell below 5 or the total sample size was \u0026lt;\u0026thinsp;40. All the statistical analyses used Python (version 3.9.12) and R software (version 4.2.1), and P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eCohort characteristics:77 patients with clinical features were divided into two groups. The most common symptoms were chest pain (n\u0026thinsp;=\u0026thinsp;47) and fever (n\u0026thinsp;=\u0026thinsp;32), followed by abdominal pain (n\u0026thinsp;=\u0026thinsp;25) and respiratory symptoms (n\u0026thinsp;=\u0026thinsp;17). A few patients developed dizziness and headache (n\u0026thinsp;=\u0026thinsp;14). 7 patients had a recent history of novel coronavirus infection (within 3 months). Most patients had abnormal ECG (77.9%), and 26 patients had ventricular tachycardia (33.8%). The elevation of the ST-segment was the second common (32.5%), followed by depression of the ST-segment (11.7%). 7 patients underwent Endomyocardial biopsy (EMB), of whom 6 were diagnosed with acute myocarditis. In addition, 6 patients had pericardial effusion. There was no significant difference in infection type, clinical manifestation, heart rate, BMI, serum biomarkers and ultrasonography result between the good and poor prognosis groups (Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). At the end of the follow-up, all patients survived, and no patients received heart transplantation.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003ePatient characteristics.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"5\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eALL\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGood prognosis\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePoor prognosis\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eP-value\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.36\u0026thinsp;\u0026plusmn;\u0026thinsp;4.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.32\u0026thinsp;\u0026plusmn;\u0026thinsp;4.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.44\u0026thinsp;\u0026plusmn;\u0026thinsp;4.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.907\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.540\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eChest pain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.304\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFever\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.781\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAbdominal pain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.807\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRespiratory symptoms\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.963\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHeadache\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.246\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eType of infection\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.055\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCRP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.74(0.89, 13.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.45(0.91,16.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.30(0.10,10.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.269\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ecTNT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e114.60(48.13,446.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e110.75(41.01,393.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e134.5(68.84,490.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.302\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMYO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e30.04(16.0,94.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e30.07(16.5,87.45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e28.01(17.0,136.56)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.800\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBNP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e254.0(90.0,626.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e206.80(72.60,445.45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e456.0(184.9,1397.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.264\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCKMB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.80(2.70,20.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.35(3.15,21.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8.9(2.55,19.15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.328\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eST elevation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u003cspan style=\"color: rgb(226, 80, 65);\"\u003e\u0026lt;\u0026thinsp;0.001\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eST depression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.812\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVentricular tachycardia\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.245\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBMI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18.21\u0026thinsp;\u0026plusmn;\u0026thinsp;3.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18.49\u0026thinsp;\u0026plusmn;\u0026thinsp;4.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17.56\u0026thinsp;\u0026plusmn;\u0026thinsp;3.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.354\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBSA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.16\u0026thinsp;\u0026plusmn;\u0026thinsp;0.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.18\u0026thinsp;\u0026plusmn;\u0026thinsp;0.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.11\u0026thinsp;\u0026plusmn;\u0026thinsp;0.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.511\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLVEDV V/B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e76.18\u0026thinsp;\u0026plusmn;\u0026thinsp;14.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e74.07\u0026thinsp;\u0026plusmn;\u0026thinsp;12.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e81.12\u0026thinsp;\u0026plusmn;\u0026thinsp;18.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.054\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLVESV V/B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e25.32(21.76,31.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e25.13(21.85,31.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e27.01(22.14,31.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.236\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLVEF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e64.0(55.0, 68.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66.0(62.0,69.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e49.0(46.0,59.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u003cspan style=\"color: rgb(226, 80, 65);\"\u003e\u0026lt;\u0026thinsp;0.001\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLVCI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.44\u0026thinsp;\u0026plusmn;\u0026thinsp;0.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLVMASS V/B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e45.70(40.14,52.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e44.89(39.59,48.74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e49.72(43.27,61.62)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRVEDV V/B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e69.12\u0026thinsp;\u0026plusmn;\u0026thinsp;17.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e69.45\u0026thinsp;\u0026plusmn;\u0026thinsp;15.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e68.35\u0026thinsp;\u0026plusmn;\u0026thinsp;20.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.796\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRVESV V/B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e28.89\u0026thinsp;\u0026plusmn;\u0026thinsp;7.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e28.84\u0026thinsp;\u0026plusmn;\u0026thinsp;6.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e29.01\u0026thinsp;\u0026plusmn;\u0026thinsp;9.57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.929\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRVEF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e57.857\u0026thinsp;\u0026plusmn;\u0026thinsp;5.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e58.74\u0026thinsp;\u0026plusmn;\u0026thinsp;4.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e55.78\u0026thinsp;\u0026plusmn;\u0026thinsp;6.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.025\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRVCI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.44\u0026thinsp;\u0026plusmn;\u0026thinsp;1.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.43\u0026thinsp;\u0026plusmn;\u0026thinsp;0.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.47\u0026thinsp;\u0026plusmn;\u0026thinsp;1.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.861\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSAXPeak Global Circumferential strain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.17(0.11,0.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.18(0.16,0.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.10(0.05,0.15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u003cspan style=\"color: rgb(226, 80, 65);\"\u003e\u0026lt;\u0026thinsp;0.001\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSAXPeak Global Radial strain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.31\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.516\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLAXPeak Global Longitudinal strain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.13\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.14\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u003cspan style=\"color: rgb(226, 80, 65);\"\u003e\u0026lt;\u0026thinsp;0.001\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLAXPeak Global Radial strain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.22\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.22\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.20\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.323\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLGE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u003cspan style=\"color: rgb(226, 80, 65);\"\u003e\u0026lt;\u0026thinsp;0.001\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\"\u003eThe data are presented as n (%) or mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD. CRP\u0026thinsp;=\u0026thinsp;C-reactive protein cTNT\u0026thinsp;=\u0026thinsp;cardiac troponinT, MYO\u0026thinsp;=\u0026thinsp;myoglobin, BNP\u0026thinsp;=\u0026thinsp;b-type natriuretic peptide, CK-MB\u0026thinsp;=\u0026thinsp;creatine kinase MB, BMI\u0026thinsp;=\u0026thinsp;body mass index, BSA\u0026thinsp;=\u0026thinsp;body surface area, LV\u0026thinsp;=\u0026thinsp;left ventricular, EDV\u0026thinsp;=\u0026thinsp;end-diastolic volume, ESV\u0026thinsp;=\u0026thinsp;end-systolic volume, EF\u0026thinsp;=\u0026thinsp;ejection fraction, CI\u0026thinsp;=\u0026thinsp;cardiac index, RV\u0026thinsp;=\u0026thinsp;right ventricular, LGE\u0026thinsp;=\u0026thinsp;late gadolinium enhancement, V/B\u0026thinsp;=\u0026thinsp;Value/Body surface area.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eCMR results: Patients with myocarditis were examined by cardiovascular MRI in the hospital within 4.5\u0026thinsp;\u0026plusmn;\u0026thinsp;7.5 days after onset. Univariate analysis revealed significant differences in LGE, LVEF, SAXGCS, LAXGLS, the LV cardiac index (LVCI), LVEF is the percentage of ejected volume from the left ventricle during systole relative to end-diastolic volume. LVMASS V/B (Value/Body surface area), RVEF, and ST elevation between the two groups, all patients with an average LVEF of 64.0%. Compared with adult myocarditis patients, LVEF damage was less common in children. The average RVEF is 57.9%. 39% of LGE cases were located in the inferior lateral wall, and 42% were in the middle interventricular septum. There was no significant difference in other cardiac function parameters, such as the LVEDV V/B or RVCI, between the patient group and the control group (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003ePredictive performance of the ML models\u003c/h2\u003e\n \u003cp\u003eAll 32 features are included in the model input, including 16 clinical features such as demographics, clinical symptoms of myocarditis, laboratory tests, ECG results, and 16 CMR parameters (Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). Among the ML models considered, the LR model (AUC\u0026thinsp;=\u0026thinsp;0.893) is superior to the RF (AUC\u0026thinsp;=\u0026thinsp;0.884), SVC (AUC\u0026thinsp;=\u0026thinsp;0.880), and XGBOOST (AUC\u0026thinsp;=\u0026thinsp;0.840) models (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e and Table \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). The Youden index was used to optimize the index and decision threshold of each model. Under this optimization threshold, the sensitivity and specificity of the LR model are 82.0% and 94.4%, respectively. Bootstrap validation with 1,000 iterations further corroborated the LR model\u0026apos;s robustness, demonstrating consistent performance superiority (AUC\u0026thinsp;=\u0026thinsp;0.895) over RF (AUC\u0026thinsp;=\u0026thinsp;0.865), SVC (AUC\u0026thinsp;=\u0026thinsp;0.862), and XGBoost ( AUC\u0026thinsp;=\u0026thinsp;0.828). The ROC curves for all models, with translucent shading representing 95% confidence intervals (CIs) superimposed, are available in the Supplementary Material 2.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003ePerformance of each predictive model.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"5\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSVC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAccuracy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.870\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.857\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.805\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.870\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSensitivity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.820\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.786\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.736\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.795\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSpecificity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.944\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.963\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.907\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.981\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAUC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.893\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.884\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.880\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.840\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePPV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.842\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.875\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.722\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.933\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNPV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.879\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.852\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.831\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.855\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eF1-score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.836\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.811\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.751\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.825\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\"\u003eDecision thresholds were optimized via the Youden index.\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\"\u003eLR\u0026thinsp;=\u0026thinsp;logistic regression, RF\u0026thinsp;=\u0026thinsp;random forest, SVC\u0026thinsp;=\u0026thinsp;support vector machine classifier, XGBoost\u0026thinsp;=\u0026thinsp;extreme gradient boosting, AUC\u0026thinsp;=\u0026thinsp;area under the curve, CI\u0026thinsp;=\u0026thinsp;confidence interval, PPV\u0026thinsp;=\u0026thinsp;positive predictive value, NPV\u0026thinsp;=\u0026thinsp;negative predictive value.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003c/div\u003e\u003cbr\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eModel interpretation\u003c/h2\u003e\n \u003cp\u003eShap software was used to explain the prediction of four ML models (LR, SVC, RF, and XGBOOST). Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e shows the functional importance ranking based on the mean | Shap Value |, with the top ten features filtered out for each model. Among all the models, LGE, LVEF, SAXGCS and LAXGLS were the four most important features. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e shows the positive or negative contributions of the top 10 features to the prognosis of myocarditis in children. In Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e, each point represents a data sample, and the color indicates whether the observation of the feature itself is greater (redder) or lower (bluer). The features, including LGE, reduced SAXGCS, impaired LAXGLS, decreased LVEF, and ST-segment elevation on ECG, demonstrate a positive correlation with ACE.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, the clinical and CMR data of children with myocarditis were involved to predict the prognosis via ML algorithm, including LR, RF, SVC, and XGBoost models. The prediction effect of the LR model was the best, and the AUC value was the highest. The four ML models screened out important prognostic factors, including LGE, LVEF, SAXGCS, LAXGLS, and ECG ST-segment elevation. Predicting the occurrence of ACE in children with myocarditis is important for early clinical treatment. This study chose LR and nonlinear models (including SVC, RF, and XGBoost) to ensure the reliability and robustness of our results. This approach helps to mitigate potential biases or inaccuracies that might arise from relying on a single model, thereby providing a more balanced and comprehensive evaluation of the data. In the receiver operating characteristic (ROC) analysis, the LR model had the highest AUC value of 0.893 and outperformed the LR, SVC, and XGBoost models.\u003c/p\u003e \u003cp\u003eLGE is a marker of myocardial fibrosis and can reflect the existence of myocardial fibrosis[8, 9]. Related studies have confirmed the value of LGE in predicting the prognosis of patients with myocarditis[10\u0026ndash;13]. It is reported that the presence of LGE is an independent predictor of poor prognosis, defined as heart transplantation, the need for extracorporeal membrane oxygenation or a ventricular assist device, and death[24, 25], which is consistent with our results. In addition, a recent long-term prognosis study in patients with acute myocarditis revealed that NYHA functional grade II and a larger range of LGE were independent predictors of long-term MACE[26]. However, Aquaro et al.[10, 11]found that LGE in patients with myocarditis mainly existed in the subepicardium of the inferior lateral wall (41%) and anterior middle septum (36%). In contrast, patients with LGE located in the anterior middle septum had a poor prognosis, similar to the results of our study. In our study, LGE was detected among 24.7% of the patients, of which approximately 39% were in the inferior lateral wall and 42% in the middle interventricular septum. Therefore, we believe that the location of LGE has greater prognostic significance than the range of LGE, which is consistent with previous studies[10, 11]. This conclusion has high clinical value and can be used to guide the clinical prediction of myocarditis and to take preventive measures as soon as possible.\u003c/p\u003e \u003cp\u003eLVEF refers to the percentage of left ventricular end-systolic ejection volume to left ventricular end-diastolic volume (LVEDV), which can reflect the degree of impaired cardiac function. Some studies have shown that the value of LVEF is lower in patients who died from myocarditis[27]. Patients with an LVEF\u0026thinsp;\u0026lt;\u0026thinsp;0.30 have a poor prognosis and a significantly increased risk of mechanical circulatory support, death, or heart transplantation (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001)[27]. Cardiac dysfunction can strongly activate the natriuretic peptide system, and increased ventricular load can lead to the release of B-type natriuretic peptide (BNP). Patients with myocarditis with an LVEF\u0026thinsp;\u0026lt;\u0026thinsp;0.50 had higher levels of BNP[28], and their peak BNP (P\u0026thinsp;\u0026gt;\u0026thinsp;10000 ng/L) was a risk factor for poor prognosis of pediatric myocarditis[25].\u003c/p\u003e \u003cp\u003eMyocardial strain (MS) refers to the degree of deformation of a cardiac segment from its original shape (end-diastole) to its maximum length (end-systole) in a specified direction and is expressed in percentage terms of the deformation. FT-CMR (feather tracking cardiovascular magnetic resonance imaging) can be used to quantitatively evaluate changes in myocardial strain of patients with myocarditis based on conventional CMR films and analyze the degree and difference in myocardial systolic and diastolic function damage, which is highly valuable for the diagnosis and prognosis of myocarditis. According to a recent review[29], a decreased GCS, or locally circumferential myocardial dysfunction, represents a response to increased wall stress and reflects local changes in myocardial characteristics, such as fibrosis or ischemia caused by microvascular disease or coronary atherosclerotic heart disease (CAD). This increased afterload may lead to progressive myocardial remodeling and the development of dysfunction, leading to a poor prognosis[30]. In addition, there was a significant correlation between the GCS and the LV quality index, which reemphasized the relationship between strain reduction and subclinical heart failure, which may be transformed into symptomatic disease due to poor ventricular remodeling[30, 31]. In addition, studies have shown that the left ventricular strain parameters, especially GLS, are impaired in patients with myocarditis compared with healthy volunteers[32, 33]. GLS is considered to be a strong predictor of major ACE in immune point inhibitor-associated myocarditis[32]. Changes in myocardial strain parameters, especially GLS, are considered to be useful in detecting early changes in cardiac insufficiency[34\u0026ndash;36]. Early cardiac dysfunction in most progressive cardiomyopathies leads to a decrease in left ventricular longitudinal mechanics, especially in dilated cardiomyopathy[37]. GLS is an independent predictor of survival in patients with dilated cardiomyopathy[34]. It is speculated that in children with myocarditis, more severe damage to the myocardial short-axis GCS and long-axis GLS strain parameters will increase the risk of left ventricular dysfunction and long-term dilated cardiomyopathy, thus affect the prognosis of patients.\u003c/p\u003e \u003cp\u003eST-segment elevation is the most common change in the ST-segment in acute myocarditis, but ST-segment depression also occurs. In myocarditis, two ST-segment elevation patterns have been described: pericarditis or the typical mode of myocardial infarction. In a study by Nucifora G et al., total ST-segment elevation in all leads was more significant in patients with larger LGE[38]. Our model shows that ECG ST-segment elevation is closely related to prognosis in children with myocarditis.\u003c/p\u003e \u003cp\u003eThis retrospective study has several limitations. First, the relatively small sample size limits the generalizability of the findings and precludes subgroup analysis. Future studies with larger cohorts and prospective designs are needed to enhance the robustness and persuasiveness of the results. Second, most patients were diagnosed clinically without endocardial biopsy confirmation, which may introduce diagnostic uncertainty. Third, the latest Lake Louise criteria have reduced the diagnostic emphasis on early gadolinium enhancement (EGE), leading to its exclusion from the final analysis. This may affect the comprehensiveness of the imaging assessment. Addressing these limitations in future research could further strengthen the validity and applicability of the findings.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn summary, ML models, particularly the LR model, which are based on clinical and imaging data from pediatric myocarditis patients, can effectively predict the prognosis of children with myocarditis. The optimal ML model (LR) offers early warning capabilities and supports more informed treatment strategies. This study will further investigate laboratory and imaging data related to pediatric myocarditis to refine the models and achieve greater diagnostic accuracy.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate:\u0026nbsp;\u003c/strong\u003eThe study was performed with approvals from the Second Affiliated Hospital of Soochow University institutional board and ethical committee, and was carried out in strict accordance with the relevant guidelines for the acquisition and use of human information and specimens, and the Declaration of Helsinki. Informed consent was waived because of the retrospective nature of the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number:\u0026nbsp;\u003c/strong\u003eNot applicable.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: Dongliang Hu, Manman Cui.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData curation: Xueke Zhang, Yuanyuan Wu.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFormal analysis: Yan Liu，Duchang Zhai.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFunding acquisition: Guohua Fan, Wu Cai.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eInvestigation: Wanliang Guo. Methodology: Dongliang Hu.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eProject administration: Guohua Fan\u0026nbsp;，Wu Cai.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eResources: Wanliang Guo，Wu Cai.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSoftware: Manman Cui.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSupervision: Shenghong Ju, Wu Cai.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eValidation: Dongliang Hu, Manman Cui.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eVisualization: Dongliang Hu, Manman Cui.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWriting\u0026mdash;original draft: Dongliang Hu, Manman Cui.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWriting\u0026mdash;review \u0026amp; editing: all authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u0026nbsp;\u003c/strong\u003eThe authors have agreed to publish this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of interest statement:\u003c/strong\u003e The authors have no conflicts of interest with respect to this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials:\u0026nbsp;\u003c/strong\u003eThe datasets generated or analyzed during the study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u0026nbsp;\u003c/strong\u003eThis study was supported by the Project of State Key Laboratory of Radiation Medicine and Protection, Soochow University (GZK1202136).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eSagar S, Liu PP, Cooper LT, Jr. Myocarditis. Lancet (London, England).2012; 379(9817):738-747. http://doi.org/10.1016/s0140-6736(11)60648-x\u003c/li\u003e\n \u003cli\u003eShiravi AA, Ardekani A, Sheikhbahaei E, Heshmat-Ghahdarijani K. Cardiovascular Complications of SARS-CoV-2 Vaccines: An Overview. Cardiology and therapy.2022; 11(1):13-21. http://doi.org/10.1007/s40119-021-00248-0\u003c/li\u003e\n \u003cli\u003eLee JW, Jeong YJ, Lee G, Lee NK, Lee HW, Kim JY, Choi BS, Choo KS. Predictive Value of Cardiac Magnetic Resonance Imaging-Derived Myocardial Strain for Poor Outcomes in Patients with Acute Myocarditis. Korean journal of radiology.2017; 18(4):643-654. http://doi.org/10.3348/kjr.2017.18.4.643\u003c/li\u003e\n \u003cli\u003eWeinreich MA, Jabbar AY, Malguria N, Haley RW. New-Onset Myocarditis in an Immunocompetent Adult with Acute Metapneumovirus Infection. Case reports in medicine.2015; 2015:814269. http://doi.org/10.1155/2015/814269\u003c/li\u003e\n \u003cli\u003eFerreira VM, Schulz-Menger J, Holmvang G, Kramer CM, Carbone I, Sechtem U, Kindermann I, Gutberlet M, Cooper LT, Liu P\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Cardiovascular Magnetic Resonance in Nonischemic\u0026nbsp;Myocardial Inflammation: Expert Recommendations. Journal of the American College of Cardiology.2018; 72(24):3158-3176. http://doi.org/10.1016/j.jacc.2018.09.072\u003c/li\u003e\n \u003cli\u003eWang H, Zhao B, Jia H, Gao F, Zhao J, Wang C. A retrospective study: cardiac MRI of fulminant myocarditis in children-can we evaluate the short-term outcomes? PeerJ.2016; 4:e2750. http://doi.org/10.7717/peerj.2750\u003c/li\u003e\n \u003cli\u003eDi Filippo S. Improving outcomes of acute myocarditis in children. Expert review of cardiovascular therapy.2016; 14(1):117-125. http://doi.org/10.1586/14779072.2016.1114884\u003c/li\u003e\n \u003cli\u003eMessroghli DR, Moon JC, Ferreira VM, Grosse-Wortmann L, He T, Kellman P, Mascherbauer J, Nezafat R, Salerno M, Schelbert EB\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: A consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR) endorsed by the European Association for Cardiovascular Imaging (EACVI). Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance.2017; 19(1):75. http://doi.org/10.1186/s12968-017-0389-8\u003c/li\u003e\n \u003cli\u003eFlorian A, Ludwig A, R\u0026ouml;sch S, Yildiz H, Sechtem U, Yilmaz A. Myocardial fibrosis imaging based on T1-mapping and extracellular volume fraction (ECV) measurement in muscular dystrophy patients: diagnostic value compared with conventional late gadolinium enhancement (LGE) imaging. European heart journal Cardiovascular Imaging.2014; 15(9):1004-1012. http://doi.org/10.1093/ehjci/jeu050\u003c/li\u003e\n \u003cli\u003eAquaro GD, Ghebru Habtemicael Y, Camastra G, Monti L, Dellegrottaglie S, Moro C, Lanzillo C, Scatteia A, Di Roma M, Pontone G\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Prognostic Value of Repeating Cardiac\u0026nbsp;Magnetic Resonance in Patients\u0026nbsp;With Acute Myocarditis. Journal of the American College of Cardiology.2019; 74(20):2439-2448. http://doi.org/10.1016/j.jacc.2019.08.1061\u003c/li\u003e\n \u003cli\u003eAquaro GD, Perfetti M, Camastra G, Monti L, Dellegrottaglie S, Moro C, Pepe A, Todiere G, Lanzillo C, Scatteia A\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Cardiac MR With Late Gadolinium Enhancement in Acute Myocarditis With\u0026nbsp;Preserved Systolic Function: ITAMY Study. Journal of the American College of Cardiology.2017; 70(16):1977-1987. http://doi.org/10.1016/j.jacc.2017.08.044\u003c/li\u003e\n \u003cli\u003eBlissett S, Chocron Y, Kovacina B, Afilalo J. Diagnostic and prognostic value of cardiac magnetic resonance in acute myocarditis: a systematic review and meta-analysis. The international journal of cardiovascular imaging.2019; 35(12):2221-2229. http://doi.org/10.1007/s10554-019-01674-x\u003c/li\u003e\n \u003cli\u003eYang F, Wang J, Li W, Xu Y, Wan K, Zeng R, Chen Y. The prognostic value of late gadolinium enhancement in myocarditis and clinically suspected myocarditis: systematic review and meta-analysis. European radiology.2020; 30(5):2616-2626. http://doi.org/10.1007/s00330-019-06643-5\u003c/li\u003e\n \u003cli\u003eLeiner T, Rueckert D, Suinesiaputra A, Bae\u0026szlig;ler B, Nezafat R, I\u0026scaron;gum I, Young AA. Machine learning in cardiovascular magnetic resonance: basic concepts and applications. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance.2019; 21(1):61. http://doi.org/10.1186/s12968-019-0575-y\u003c/li\u003e\n \u003cli\u003eZhang N, Yang G, Gao Z, Xu C, Zhang Y, Shi R, Keegan J, Xu L, Zhang H, Fan Z\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Deep Learning for Diagnosis of Chronic Myocardial Infarction on Nonenhanced Cardiac Cine MRI. Radiology.2019; 291(3):606-617. http://doi.org/10.1148/radiol.2019182304\u003c/li\u003e\n \u003cli\u003eBaessler B, Mannil M, Oebel S, Maintz D, Alkadhi H, Manka R. Subacute and Chronic Left Ventricular Myocardial Scar: Accuracy of Texture Analysis on Nonenhanced Cine MR Images. Radiology.2018; 286(1):103-112. http://doi.org/10.1148/radiol.2017170213\u003c/li\u003e\n \u003cli\u003eLaw YM, Lal AK, Chen S, Čih\u0026aacute;kov\u0026aacute; D, Cooper LT, Jr., Deshpande S, Godown J, Grosse-Wortmann L, Robinson JD, Towbin JA. Diagnosis and Management of Myocarditis in Children: A Scientific Statement From the American Heart Association. Circulation.2021; 144(6):e123-e135. http://doi.org/10.1161/cir.0000000000001001\u003c/li\u003e\n \u003cli\u003eHsiao JF, Koshino Y, Bonnichsen CR, Yu Y, Miller FA, Jr., Pellikka PA, Cooper LT, Jr., Villarraga HR. Speckle tracking echocardiography in acute myocarditis. The international journal of cardiovascular imaging.2013; 29(2):275-284. http://doi.org/10.1007/s10554-012-0085-6\u003c/li\u003e\n \u003cli\u003eFriedrich MG, Sechtem U, Schulz-Menger J, Holmvang G, Alakija P, Cooper LT, White JA, Abdel-Aty H, Gutberlet M, Prasad S\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Cardiovascular magnetic resonance in myocarditis: A JACC White Paper. Journal of the American College of Cardiology.2009; 53(17):1475-1487. http://doi.org/10.1016/j.jacc.2009.02.007\u003c/li\u003e\n \u003cli\u003eSchultz JC, Hilliard AA, Cooper LT, Jr., Rihal CS. Diagnosis and treatment of viral myocarditis. Mayo Clinic proceedings.2009; 84(11):1001-1009. http://doi.org/10.1016/s0025-6196(11)60670-8\u003c/li\u003e\n \u003cli\u003eMahrholdt H, Goedecke C, Wagner A, Meinhardt G, Athanasiadis A, Vogelsberg H, Fritz P, Klingel K, Kandolf R, Sechtem U. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation.2004; 109(10):1250-1258. http://doi.org/10.1161/01.Cir.0000118493.13323.81\u003c/li\u003e\n \u003cli\u003eCaforio AL, Pankuweit S, Arbustini E, Basso C, Gimeno-Blanes J, Felix SB, Fu M, Heli\u0026ouml; T, Heymans S, Jahns R\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: a position statement of the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. European heart journal.2013; 34(33):2636-2648, 2648a-2648d. http://doi.org/10.1093/eurheartj/eht210\u003c/li\u003e\n \u003cli\u003eMagnani JW, Dec GW. Myocarditis: current trends in diagnosis and treatment. Circulation.2006; 113(6):876-890. http://doi.org/10.1161/circulationaha.105.584532\u003c/li\u003e\n \u003cli\u003eGr\u0026uuml;n S, Schumm J, Greulich S, Wagner A, Schneider S, Bruder O, Kispert EM, Hill S, Ong P, Klingel K\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Long-term follow-up of biopsy-proven viral myocarditis: predictors of mortality and incomplete recovery. Journal of the American College of Cardiology.2012; 59(18):1604-1615. http://doi.org/10.1016/j.jacc.2012.01.007\u003c/li\u003e\n \u003cli\u003eSachdeva S, Song X, Dham N, Heath DM, DeBiasi RL. Analysis of clinical parameters and cardiac magnetic resonance imaging as predictors of outcome in pediatric myocarditis. The American journal of cardiology.2015; 115(4):499-504. http://doi.org/10.1016/j.amjcard.2014.11.029\u003c/li\u003e\n \u003cli\u003eAndr\u0026eacute; F, Stock FT, Riffel J, Giannitsis E, Steen H, Scharhag J, Katus HA, Buss SJ. Incremental value of cardiac deformation analysis in acute myocarditis: a cardiovascular magnetic resonance imaging study. The international journal of cardiovascular imaging.2016; 32(7):1093-1101. http://doi.org/10.1007/s10554-016-0878-0\u003c/li\u003e\n \u003cli\u003eSchubert S, Opgen-Rhein B, Boehne M, Weigelt A, Wagner R, M\u0026uuml;ller G, Rentzsch A, Zu Knyphausen E, Fischer M, Papakostas K\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Severe heart failure and the need for mechanical circulatory support and heart transplantation in pediatric patients with myocarditis: Results from the prospective multicenter registry \u0026quot;MYKKE\u0026quot;. Pediatric transplantation.2019; 23(7):e13548. http://doi.org/10.1111/petr.13548\u003c/li\u003e\n \u003cli\u003eAkg\u0026uuml;l F, Er A, Ulusoy E, \u0026Ccedil;ağlar A, Vuran G, Seven P, Yılmazer MM, Ağın H, Apa H. Are clinical features and cardiac biomarkers at admission related to severity in pediatric acute myocarditis?: Clinical features and cardiac biomarkers in pediatric acute myocarditis. Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.2022; 29(5):376-380. http://doi.org/10.1016/j.arcped.2022.03.008\u003c/li\u003e\n \u003cli\u003eKorosoglou G, Sagris M, Andr\u0026eacute; F, Steen H, Montenbruck M, Frey N, Kelle S. Systematic review and meta-analysis for the value of cardiac magnetic resonance strain to predict cardiac outcomes. Scientific reports.2024; 14(1):1094. http://doi.org/10.1038/s41598-023-50835-5\u003c/li\u003e\n \u003cli\u003eKass DA. Ventricular arterial stiffening: integrating the pathophysiology. (1524-4563 (Electronic)).\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eRosen BD, Edvardsen T Fau - Lai S, Lai S Fau - Castillo E, Castillo E Fau - Pan L, Pan L Fau - Jerosch-Herold M, Jerosch-Herold M Fau - Sinha S, Sinha S Fau - Kronmal R, Kronmal R Fau - Arnett D, Arnett D Fau - Crouse JR, 3rd, Crouse Jr 3rd Fau - Heckbert SR\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Left ventricular concentric remodeling is associated with decreased global and regional systolic function: the Multi-Ethnic Study of Atherosclerosis. (1524-4539 (Electronic)).\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eAwadalla M, Mahmood SS, Groarke JD, Hassan MZO, Nohria A, Rokicki A, Murphy SP, Mercaldo ND, Zhang L, Zlotoff DA\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Global Longitudinal Strain and Cardiac Events in Patients With Immune Checkpoint Inhibitor-Related Myocarditis. Journal of the American College of Cardiology.2020; 75(5):467-478. http://doi.org/10.1016/j.jacc.2019.11.049\u003c/li\u003e\n \u003cli\u003eLuetkens JA, Schlesinger-Irsch U, Kuetting DL, Dabir D, Homsi R, Doerner J, Schmeel FC, Fimmers R, Sprinkart AM, Naehle CP\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Feature-tracking myocardial strain analysis in acute myocarditis: diagnostic value and association with myocardial oedema. European radiology.2017; 27(11):4661-4671. http://doi.org/10.1007/s00330-017-4854-4\u003c/li\u003e\n \u003cli\u003eBuss SJ, Breuninger K, Lehrke S, Voss A, Galuschky C, Lossnitzer D, Andre F, Ehlermann P, Franke J, Taeger T\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e. Assessment of myocardial deformation with cardiac magnetic resonance strain imaging improves risk stratification in patients with dilated cardiomyopathy. European heart journal Cardiovascular Imaging.2015; 16(3):307-315. http://doi.org/10.1093/ehjci/jeu181\u003c/li\u003e\n \u003cli\u003ePedrizzetti G, Claus P, Kilner PJ, Nagel E. Principles of cardiovascular magnetic resonance feature tracking and echocardiographic speckle tracking for informed clinical use. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance.2016; 18(1):51. http://doi.org/10.1186/s12968-016-0269-7\u003c/li\u003e\n \u003cli\u003eAmzulescu MS, De Craene M, Langet H, Pasquet A, Vancraeynest D, Pouleur AC, Vanoverschelde JL, Gerber BL. Myocardial strain imaging: review of general principles, validation, and sources of discrepancies. European heart journal Cardiovascular Imaging.2019; 20(6):605-619. http://doi.org/10.1093/ehjci/jez041\u003c/li\u003e\n \u003cli\u003e\u0026nbsp; Claus P, Omar AMS, Pedrizzetti G, Sengupta PP, Nagel E. Tissue Tracking Technology for Assessing Cardiac Mechanics: Principles, Normal Values, and Clinical Applications. JACC Cardiovascular imaging.2015; 8(12):1444-1460. http://doi.org/10.1016/j.jcmg.2015.11.001\u003c/li\u003e\n \u003cli\u003e\u0026nbsp; Nucifora G, Miani D, Di Chiara A, Piccoli G, Artico J, Puppato M, Slavich G, De Biasio M, Gasparini D, Proclemer A. Infarct-like acute myocarditis: relation between electrocardiographic findings and myocardial damage as assessed by cardiac magnetic resonance imaging. Clinical cardiology.2013; 36(3):146-152. http://doi.org/10.1002/clc.22088\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-pediatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bped","sideBox":"Learn more about [BMC Pediatrics](http://bmcpediatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bped/default.aspx","title":"BMC Pediatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Myocarditis, magnetic resonance imaging, major cardiovascular adverse events, machine learning","lastPublishedDoi":"10.21203/rs.3.rs-5534455/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5534455/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eTo develop machine learning (ML) models incorporating explanatory cardiac magnetic resonance (CMR) parameters for predicting the prognosis of myocarditis in pediatric patients.\u003c/p\u003e\u003ch2\u003eMaterials and Methods\u003c/h2\u003e \u003cp\u003e77 patients with pediatric myocarditis diagnosed clinically between January 2020 and December 2023 were enrolled retrospectively. All patients were examined by ultrasound, electrocardiogram (ECG), serum biomarkers on admission, and CMR scan to obtain 16 explanatory CMR parameters. All patients underwent follow-up echocardiography and CMR. Patients were divided into two groups according to the occurrence of adverse cardiac events (ACE) during follow-up: the poor prognosis group (n\u0026thinsp;=\u0026thinsp;23) and the good prognosis group (n\u0026thinsp;=\u0026thinsp;54). Four models were established, including logistic regression (LR), random forest (RF), support vector machine classifier (SVC), and extreme gradient boosting (XGBoost) model. The performance of each model was evaluated by the area under the receiver operating characteristic curve (AUC). Model interpretation was generated by Shapley additive interpretation.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eAmong the four models, the three most important features were late gadolinium enhancement (LGE), left ventricular ejection fraction (LVEF), and SAXPeak Global Circumferential Strain (SAXGCS). In addition, LGE, LVEF, SAXGCS, and LAXPeak Global Longitudinal Strain (LAXGLS) were selected as the key predictors for all four models. Four interpretable CMR parameters were extracted, among which the LR model had the best prediction performance. The AUC, sensitivity, and specificity were 0.893, 0.820, and 0.944, respectively. The findings indicate that the presence of LGE on CMR imaging, along with reductions in LVEF, SAXGCS, and LAXGLS, are predictive of poor prognosis in patients with acute myocarditis.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eML models, particularly the LR model, demonstrate the potential to predict the prognosis of children with myocarditis. These findings provide valuable insights for cardiologists, supporting more informed clinical decision-making and potentially enhancing patient outcomes in pediatric myocarditis cases.\u003c/p\u003e","manuscriptTitle":"Using machine learning models based on cardiac magnetic resonance parameters to predict the prognostic in children with myocarditis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-28 05:32:43","doi":"10.21203/rs.3.rs-5534455/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-05-02T07:44:31+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-05-02T07:43:11+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-29T00:20:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"162663753666822325270996579477053534731","date":"2025-04-22T21:39:28+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-22T13:19:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"264872167670474893077263462537348074884","date":"2025-04-22T13:15:55+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-04-22T06:35:35+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-18T06:19:28+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Pediatrics","date":"2025-04-16T09:06:38+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-pediatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bped","sideBox":"Learn more about [BMC Pediatrics](http://bmcpediatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bped/default.aspx","title":"BMC Pediatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b4e894bd-3fa2-4dac-8c6f-11bd8c0a1392","owner":[],"postedDate":"April 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-05-26T16:00:04+00:00","versionOfRecord":{"articleIdentity":"rs-5534455","link":"https://doi.org/10.1186/s12887-025-05753-y","journal":{"identity":"bmc-pediatrics","isVorOnly":false,"title":"BMC Pediatrics"},"publishedOn":"2025-05-24 15:57:12","publishedOnDateReadable":"May 24th, 2025"},"versionCreatedAt":"2025-04-28 05:32:43","video":"","vorDoi":"10.1186/s12887-025-05753-y","vorDoiUrl":"https://doi.org/10.1186/s12887-025-05753-y","workflowStages":[]},"version":"v1","identity":"rs-5534455","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5534455","identity":"rs-5534455","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}